The last decade has provided several proposals on pressure containment criteria for pipelines, also reflected in recent code updates, e.g. ISO, DNV OS-F101, Germaniseher Lloyd and API. This paper will examines two aspects often discussed in papers in relation to these, namely what is "correct" and what is "best". From a probabilistic point of view, this determines the selection of characteristic resistance, characteristic load effect and partial safety factors. Surprisingly, design criteria are not normally established on scientific basis only but to a large extent also on engineering judgement and public opinion in order to be accepted by the industry and authorities. Hence, what is correct is also a matter of timing. This implies that the industry may see even more modifications of the existing criteria as the society becomes more mature. The paper finally indicates such foreseen future modifications.
During the last decade, the pipeline industry has slowly transformed from the classical Allowable Stress Design (ASD) format to limit state based design, normally expressed in the format of the Load and Resistance Factor Design (LRFD). In the transformation to limit state based design, the hoop stress criterion has become a criterion for pressure containment only. Traditionally, this has not been the case, at least not in the near platform zone where a low usage factor has implicitly covered other failure modes. Being now a criterion for pressure containment, it should be reassessed in order to be representative for the pipeline pressure containment capacity and an appropriate safety factor. In the SUPERB project, Jiao (1995), the pressure containment criterion was re-assessed and new usage factors determined by means of structural reliability methods. This was included with some modifications in DNV96. This basis was re-assessed including more data as part of the new revision, DNV OS-F101.
